My research mainly focus on reinforcement learning with neural networks (also called deep reinforcement learning).
We aimed at designing agents that take decisions in an unknown environment, and learn through their own interaction with this environment to maximize a given criterion.
More precisely, I am working on model-free actor-critic algorithms to deal with continuous environments (in state and action), trying to make them more data-efficient without losing the scalability of neural networks.
Keywords: Reinforcement Learning, Neural Networks, Transfer Learning, Developmental Learning, Machine Learning, Actor-Critic
||| Matthieu Zimmer and Paul Weng. Exploiting the sign of the advantage function to learn deterministic policies in continuous domains. In International Joint Conferences on Artificial Intelligence, August 2019.
[ bib | poster | hal pdf | source code | slides ]
||| Matthieu Zimmer, Yann Boniface, and Alain Dutech. Developmental reinforcement learning through sensorimotor space enlargement. In The 8th Joint IEEE International Conference on Development and Learning and on Epigenetic Robotics, September 2018.
[ bib | video | hal pdf | source code | slides ]
||| Matthieu Zimmer and Stephane Doncieux. Bootstrapping q-learning for robotics from neuro-evolution results. IEEE Transactions on Cognitive and Developmental Systems, 2017.
[ bib | video | hal pdf | pdf ]